An MLP neural network for time delay prediction in networked control systems

This paper considers networked control systems with time varying delays. The main idea is to apply a variable sampling period in order to compensate for the time delay. A feedforward multilayered neural network is first properly developed to estimate the time delay at each sampling period. Then, this predicted time delay is taken as the sampling period between the current and the next sampling steps. The simulation results show that the proposed approach makes the controller more robust for stabilizing the system by reducing remarkably the influence of the closed loop time delay.